地铁隧道沉降组合预测模型

Combined Prediction Model of Subway Tunnel Settlement

为了保障地铁营运的安全性,地铁隧道沉降监测与预测尤为必要。在传统变形预测模型的基础上,结合EMD方法在信号分解与重构中的优势,提出一种基于EMD方法的组合预测模型。首先,使用能够对数据序列进行自适应分解的EMD方法将数据序列分解为若干个不同频率的IMF分量;其次,使用GM(1,1)-小波神经网络对低频分量进行预测,使用AR模型对高频分量进行预测;最后,重构不同分量预测结果,得到最终预测结果。通过地铁沉降预测实例对组合模型、GM(1,1)模型和GM(1,1)-小波神经网络模型的预测结果进行比较,结果表明,组合模型的预测精度更高,验证了组合模型的有效性。本文提出的方法在实际工程中具有一定的应用价值。

It is particularly necessary to monitor and predict of subway tunnel settlement in order to ensure the safety of subway operation.Based on the traditional deformation prediction model and the advantages of EMD method in signal decomposition and reconstruction,a combined prediction model based on EMD method is proposed in this paper.Firstly,the EMD method,which can decompose the data sequence adaptively,is used to decompose the data sequence into several IMF components with different frequencies;Secondly,GM(1,1)wavelet neural network is used to predict the low frequency components,and AR model is used to predict the high frequency components;Finally,the prediction results of different components are reconstructed to get the final prediction results.Through the example of subway settlement prediction,the prediction results of the combined model are compared with GM(1,1)model and GM(1,1)wavelet neural network model.The results show that the prediction accuracy of the combined model is higher,which verifies the effectiveness of the combined model.The method proposed in this paper has certain application value in practical engineering.